WHO Malaria Vaccine ‘Rollout’ In Africa Raises Ethical Questions
By Deborah Borfitz
March 26, 2020 | The large-scale Malaria Vaccine Implementation Programme (MVIP) of the World Health Organization (WHO) has ignited debate within the bioethics community about whether it is really a cluster randomization study that has seriously breached international ethical standards by failing to obtain informed consent from the parents of children opting for vaccination.
As reported on the WHO website, the MVIP is providing RTS,S/AS01 (Mosquirix), a malaria vaccine developed by GlaxoSmithKline (GSK), as part of the routine immunization services of the ministries of health (MoH) of Ghana, Kenya and Malawi. The vaccine was positively reviewed by the European Medicines Agency (EMA), which provides scientific opinions on medicines that are not intended for use in the European Union but are needed to prevent or treat diseases of major public health importance around the world.
But its use is being limited to pilot implementation—in part to evaluate outstanding safety concerns that emerged from previous clinical trials.
Regions that participated in the phase III trials by GSK met one of the selection criteria for the implementation program, according to WHO. The vaccine was then allocated to randomly selected communities in each of the final three countries, a decision reached in conjunction with national authorities as “a fair way to distribute limited vaccine doses in the first part of a phased, sub-national introduction.”
GSK is donating up to 10 million doses of Mosquirix for use in the pilot, WHO reports.
The vaccine’s introduction is under the control and direction of the countries’ MoH, says a spokesperson for WHO. “The fact that randomization is being used in the delivery of the vaccine does not mean that the deployment of the vaccine through the routine system can be equated to a trial.”
On its website and in a letter of rebuttal to the ethics charge (doi: 10.1136/bmj.m734) in the British Medical Journal (BMJ), WHO maintains that the “systematic evaluation” of the MVIP rollout is good practice and not a research activity.
The letter highlights the fact that Mosquirix has been approved and therefore not an experimental vaccine, notes Jonathan Kimmelman, Ph.D., director of the biomedical ethics unit at McGill University in Montreal. That argument is a “red herring,” he adds. “The question is not whether a vaccine is experimental, but whether the evaluation it is conducting is research. We do clinical trials all the time on ‘approved’ treatments. But when we do so, we still get informed consent from research subjects.”
Just yesterday, members of the committee who wrote the human research ethics guidelines of the Council for International Organizations of Medical Sciences—the standards WHO says guides its Research Ethics Review Committee—published a letter in the BMJ that aligns with the views of Kimmelman and his colleague Charles Weijer, a bioethicist at Western University in Ontario who strongly believes that this was a malaria vaccine study in which informed consent failed to happen.
Applying ethical norms for research is “especially important in the context of implementation programs that pilot vaccines, given well-known concerns about vaccine hesitancy,” they state. Similar ethical controversy derailed rollout plans for a human papillomavirus vaccine in India, they note.
Research Or Care?
From Kimmelman’s perspective, WHO is piggybacking research atop a public health intervention. That tends to blur the line between research and routine care, and raises ethical concerns, he says.
“Calling something research entails more demanding standards oversight and informed consent,” Kimmelman explains, so “individuals conducting the evaluation may be subconsciously motivated to view an undertaking through the lens of practice rather than research.”
An estimated 720,000 children will be receiving Mosquirix, the world's first licensed malaria vaccine, over the next two years. Some districts and sub-counties within the three pilot implementation areas were randomized to receive the vaccine sooner than others, WHO says, to help understand the public health usefulness of the vaccine and determine if the vaccine should be introduced more broadly. The program began last year and is expected to continue through 2023.
According to WHO, consent for vaccinating children living in areas randomized to receive the new vaccine is “implied.” Information on vaccination is provided to the community and parents through methods such as health talks and community outreach, it says, and parents who present for vaccination “do so with the option to vaccinate their children or not.”
Waivers of informed consent are allowable under certain circumstances, says Kimmelman, and internationally the conditions typically include occasions when obtaining consent is utterly infeasible and the risks of participation are minimal. “But to me the crucial question is whether [the malaria vaccine] rollout effort was guided or informed by research objectives. My understanding is that most rollouts don’t typically use randomization to decide which jurisdictions are going to get a vaccine first and which ones get it next.”
The rollout was informed in part by a joint technical expert group (JTEG) on malaria vaccines that recommended a phased introduction of Mosquirix in three to five distinct settings with moderate to high malaria transmission, says a WHO spokesperson. WHO jointly decided with the countries’ respective ministries of health to randomize.
If WHO can elicit evidence that its rollout plan, including randomization and pacing, was “driven solely by public health objectives rather than research then I am less concerned about the waiver of informed consent,” says Kimmelman. “I’d be surprised if that were the case, however, as a proper evaluation would entail adjustment of the rollout to maximize the knowledge acquired from the process.”
It’s likely that at least some aspects of how the program is being rolled out were based on what would maximize the statistical power to detect effects, says Kimmelman, which would make this research. “We’re talking about a setting where an intervention is being tested in a population to … determine whether it’s worth expanding application of that intervention—and that’s one the of reasons I believe the standards that would apply to research ought to be applicable [here].”
People should know when they’re being enrolled in research and give their consent because research and care are “in tension with each other,” Kimmelman says. “In this case, the reason tension arises is that the particular villages [in the pilot implementation] are determined not on the basis of public health priorities but to try to answer an important question about the safety and efficacy of the vaccine.”
In the course of this investigation, he continues, “imagine if some of the safety issues they’re trying to reduce concern about [e.g., meningitis] turn out to be real safety issues, and people who are early recipients of vaccine have higher rates of those adverse events than people getting it later. If you were a patient and find out your village was selected in part to make that evaluation, you would feel like you should have been told.”
Failure to obtain informed consent from parents whose children are taking part in the study violates the Ottawa Statement, a consensus statement on the ethics of cluster randomized trials, as well as ethical guidelines of the Council for International Organizations of Medical Sciences with whom WHO collaborates. As Peter Doshi points out in the original BMJ article, human rights provisions of the Malawi constitution specifically stipulate, “No person shall be subjected to medical or scientific experimentation without his or her consent.”
Kimmelman’s “charitable view” of the apparent breach is that the ethics surrounding these types of trials is complicated and confusing in the best of times. It is also a relatively new area of research.
“People who do the research may have intuitions that are conflicting with the cutting-edge thinking of people who work in ethics,” he says, meaning entities like WHO could “easily blunder into bad decisions.” On the other hand, “we’ve now had seven years of consensus that you need to get informed consent for this particular trial model. Using cluster randomization doesn’t give you carte blanche to waive informed consent.”
In its BMJ letter, WHO states that “evaluation of the pilot implementation is being conducted in accordance with established and recognized national and international ethical standards and with respect to human subjects regulations.” Four ethical review boards—the one at WHO and in each of the three participating countries—reviewed and approved the evaluation protocols.
The large, phase III clinical trial of Mosquirix was completed in 2014 and “showed that the vaccine can significantly reduce malaria, has an acceptable safety profile and, if implemented broadly, could save tens of thousands of lives,” says a WHO spokesperson.
Safety signals that arose during phase III clinical testing of Mosquirix included more cases of meningitis in children who receive the vaccine than in those who did not, as well as more cases of cerebral malaria among a subset of participants who developed severe malaria despite the vaccine, according to WHO. “There was no evidence that the safety signals were caused by the vaccine and the EMA considered them likely chance findings.”
A post-hoc analysis also identified an imbalance in female mortality, but the EMA concluded that there was insufficient information to formally classify the finding as a potential risk, WHO says. Instead, it recommended that the concern be monitored along with the other two safety signals during vaccine introduction.
Three of the 14 secondary outcomes being evaluated during the vaccine rollout are gender-related, as reported in the study’s registration on cliniclatrials.gov—number of deaths in children by gender, number of deaths in hospitalized children by gender and number of malaria associated deaths in hospitalized children by gender.
In the training information WHO says it has shared with country partners about Mosquirix's potential risks, increased risk of death among girls is not mentioned. But if the analysis plan in the study protocol looks at that specific question, it was clearly a hypothesis that researchers were trying to address and had some credible evidence behind it, says Kimmelman.
A WHO spokesperson says “the safety signals that arose during the clinical testing of the vaccine are being carefully monitored during its introduction through surveillance and evaluation studies, as outlined in the evaluation protocols and as described in the statistical analysis plans. This includes the finding from the post-hoc analysis which identified an imbalance in female mortality in the phase III trial.”
The hypothesis related to female mortality has limited support, but the magnitude of the impact if it is proven true is such that it warrants testing, observes Kimmelman. “I think the challenge is finding an effective way to communicate that without overly reinforcing the credibility of the hypothesis while still being transparent about the fact that this is a concern that the investigation is aimed at ruling out.”
Phase IV Component
As part of the MVIP, GSK is conducting several Phase IV studies in parts of the pilot areas, according to information posted on WHO’s website. These studies are part of the risk management plan with the EMA, says a spokesperson for the agency, and “will provide additional information on the vaccine’s effectiveness and safety in routine use. This is a requirement for any new vaccine.” The data collected will complement data from the pilot evaluations led by WHO.
WHO says it has “well-established evaluation partners” from each of the three countries evaluating the pilot introduction. “The study protocols for the evaluation components—sentinel hospital, mortality surveillance and a series of household surveys—have been reviewed by the ethics review committee at WHO and by the institutional or national ethical review committees in each pilot country.”
Each of the protocols describes how the countries’ Expanded Programme on Immunizations (EPI) would introduce the vaccine, including the random selection of areas to introduce the vaccine first, and have been reviewed by national regulatory authorities in each of the three implementing countries, says a WHO spokesperson. The consent process for the routine administration of the malaria vaccine is the same “opt-out approach” (aka “implied consent”) used for all vaccines provided through EPI.
Informed consent is being obtained prior to collecting data from participants in the evaluations that are beyond routine care, according to a WHO spokesperson. “In Malawi and Kenya, the IRB provided a waiver of consent for the collection of [anonymized] routine hospital surveillance data.”
WHO’s evaluation is registered as an observational study evaluating “the impact and feasibility of delivering the four recommended doses of the vaccine and consolidation of its safety profile,” the agency reports.
Ethical debates like this are not completely new, notes Kimmelman. Many previous trials have been mired in controversy around the question of informed consent because of the “incredibly fuzzy” line between research and care.
In 2007, the Office of Human Research Protection determined that Johns Hopkins IRB acted inappropriately by categorizing a Keystone ICU quality improvement initiative as “exempt” research not requiring individual, written informed consent. The project involved implementation of a checklist to improve the culture of safety in hospitals and encourage physicians to follow evidence-based practices when inserting venous catheters.
After results of a rare donor-intervention trial published in 2015, the IRB at the University of California San Francisco entered the spotlight for its decision not to require transplant doctors to get informed consent from donors (reasoning that they were dead and thus not human research subjects) and recipients (reasoning that the hypothermia protocol posed minimal risk to them), says Kimmelman. Opinions were mixed on whether the donor or recipient’s hospital IRB should have reviewed the trial, who is a human research subject and what kind of consent is necessary.
The IRB at the University of Alabama in Birmingham's was alternately defended and scolded for allowing the Surfactant, Positive Pressure, and Oxygenation Randomized Trial (SUPPORT) to proceed without informing parents of the risks of death and blindness to their premature infant, he says. The study was trying to determine the optimum oxygen saturation level to improve survival.
Editor's Note: Attribution for the WHO quotes has been changed to reflect that the responses included input from several WHO representatives.